O'Reilly logo

Mobile Agents by Wilhelm R. Rossak, Peter Braun

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

132 Chapter 4 Mobile Agent Communication
4.1 Introduction
Problems related to agent communication have been extensively discussed
in the area of distributed intelligence and multi-agent systems. However, it
should be obvious that communication is also one of the most important
challenges in mobile applications. Applications beyond a specific level of
complexity should not be built using only a single agent that carries all knowl-
edge and all strategies. Rather, it would be better to model a world of different
agents, where each one is specialized to solve a specific problem [Glaser,
2002].
Let’s look at a simple example in which an agent will arrange a business
trip for a human user. The business trip consists of booking a flight and a
hotel, renting a car, arranging a meeting with the business partner, and later
checking in with the airline if the flight is delayed or canceled. In this case the
agent should react to the situation appropriately by informing the business
partner and the hotel.
A possible architecture for this application might consist of several spe-
cialized agents to find information regarding flights, other agents that are
responsible for finding an adequate hotel, and so forth. These are the slave
agents, which are instructed by a single master agent that communicates to
the user and monitors the entire process. Obviously, all these agents must
exchange information about their tasks and intermediate results. Not only
does the master instruct slave agents, it also sometimes terminates slave
agents or changes some parameters in response to a change in the user
requirements. The main problem in enabling mobile agents to communicate
with each other is to locate agents that can move through the network with-
out knowing their current location in advance. In addition, the slave agents
might want to periodically check if their master agent is still alive, thus per-
forming some kind of orphan detection that also might require locating the
master if it is allowed to be mobile too.
In fact, agent communication includes several issues—most of them
beyond the scope of this book. (They are common in all multi-agent systems
and have therefore already been discussed in other, more specialized books.)
For example, the problem of how to design such multi-agent systems affects
parts of software engineering [Plekhanova, 2002], and the problem of design-
ing suitable communication paradigms for agents is part of research that is
done in distributed artificial intelligence [Ferber, 1999]. Also, the definition
of languages that are appropriate to communicate between agents is part
of existing research in the area of multi-agent systems and is not a problem

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required